Orientation
What You'll Master Here
Most people meet data modeling as a pile of disconnected jargon: normalization, star schemas, surrogate keys, SCD type 2. This chapter steps back and gives you the one frame that makes all of it make sense, so the rest of the course clicks into place instead of feeling like trivia.
You will learn what a data model actually is, why a deliberately designed model is worth real money (and a bad one is a slow, expensive tax), and the single framework every modeler uses: the conceptual, logical, and physical levels.
Crucially, this is not theory for its own sake. By the end you should be able to look at a vague business request and know the first questions to ask before you ever type CREATE TABLE. That is the gap between "knowing the words" and "being able to build the thing."
Why it matters
Every later chapter, from normalization to dimensional modeling to data contracts, is a specific tool. This chapter is the toolbox: the shared way of thinking that tells you which tool to reach for and why.
Core mental model
A data model is a set of deliberate decisions about how to represent the real world in data, captured at three levels of detail.
- data model
- A deliberate structure for data: the entities, attributes, keys, and relationships you choose to represent.
- schema
- The concrete shape of data in a system (tables, columns, types); the physical end of a model.
- entity
- A thing the business cares about and stores data about: customer, order, product.
- relationship
- How entities connect: a customer places orders, an order contains items.
Common mistake
Treating data modeling as drawing boxes and lines after the fact.
The model becomes documentation nobody trusts, while the real structure grows by accident in the database.
Better habit
- Treat the model as a set of decisions, not a diagram.
- Ask what questions the data must answer before designing it.
- Use the three levels as a checklist on every modeling task.
Data modeling is decision-making, not drawing. The diagram is just a record of decisions about grain, keys, relationships, and history. Get the decisions right and the diagram writes itself.
Each topic on the left adds one layer. Read them in order; the worked example in the middle ties conceptual, logical, and physical together on one real schema.
Practice prompts
- Write one sentence defining a data model in your own words.
- Name three entities and one relationship from an app you use daily.
Remember this
Data modeling is the craft of deciding how to represent reality in data; this chapter gives you the frame that the whole course builds on.